Update app.py

app.py

@@ -3,28 +3,43 @@ from tensorflow.keras.models import load_model
 from tensorflow.keras.preprocessing.image import img_to_array
 import numpy as np
 from PIL import Image
+import matplotlib.pyplot as plt
 
-# Load model
+# Load your trained model
 model = load_model("waste_classification(Mobilenetv2).h5")
 class_names = ['cardboard', 'glass', 'metal', 'paper', 'plastic', 'trash']
 
-# Prediction function
-def predict_from_webcam(image):
+# Prediction function: returns label + chart
+def predict_with_chart(image):
     image = image.resize((224, 224))
     img_array = img_to_array(image) / 255.0
     img_array = np.expand_dims(img_array, axis=0)
 
     prediction = model.predict(img_array)[0]
-    predicted_class = class_names[np.argmax(prediction)]
+    pred_index = np.argmax(prediction)
+    pred_label = class_names[pred_index]
     confidence = float(np.max(prediction))
 
-    return f"{predicted_class} ({confidence*100:.1f}%)"
+    # Plot bar chart
+    fig, ax = plt.subplots(figsize=(6, 4))
+    ax.bar(class_names, prediction, color='skyblue')
+    ax.set_ylabel('Probability')
+    ax.set_ylim(0, 1)
+    ax.set_title('Class Probabilities')
+    plt.xticks(rotation=45)
+    plt.tight_layout()
 
-# Use gr.Camera() for webcam input (Gradio 4.x)
+    return f"Prediction: {pred_label} ({confidence*100:.1f}%)", fig
+
+# Gradio interface: webcam + upload
 gr.Interface(
-    fn=predict_from_webcam,
-    inputs=gr.Camera(type="pil"),  # webcam input as PIL image
-    outputs="text",
-    title="Live Waste Classifier",
-    description="Point your webcam at waste material. The model will classify it live."
+    fn=predict_with_chart,
+    inputs=gr.Image(type="pil", label="Upload or Capture Waste Image", source="upload"),
+    outputs=[
+        gr.Textbox(label="Predicted Class"),
+        gr.Plot(label="Class Probability Chart")
+    ],
+    title="Waste Classifier - Upload or Webcam",
+    description="Upload a waste image or capture using webcam. The model shows predicted class and probability chart.",
+    live=True
 ).launch()